CHAPTER 4

IMPROVEMENTS TO BITUMEN EMULSIONS

4.1 Introduction

With the development of bituminous emulsion in recent past, the bitumen and
cutback bitumen used as the binder have partially replaced by bitumen emulsion.
Percentage of use of bitumen emulsion has been gradually increased with time and
the production units have also been improved for the supply of the demand. At
present almost all the surface treatment is done by using bitumen emulsion as the
binder in Sri Lanka.

Surface dressing done with bitumen emulsion had been locally modified over many
years to provide a fast and economical means of increasing the life of local roads
and footways. However it was difficult to achieve consistent in performance.
Stripping of aggregate, flushing of the surface, peeling off patches, cracks and
streaking along the surface were some common problems and stripping of
aggregate was the most predominate out of them when emulsion was used.

As a solution it was suggested by RDA to apply a sand seal over the surface
dressing, DBST and SBST in order to overcome the problem of stripping of
aggregate. However this option was costly and may not be an appropriate solution.

Therefore a research work was commenced in order to find the reasons for frequent
premature failure of surface dressings and propose sustainable solutions to
minimize failures. Results of the tests carried out on the samples of Cationic Rapid
Setting-2 (CRS-2) emulsion sent by the Plant Manager, Emulsion plant, RC&DC to
R&D division of RDA and tested by the staff attached to R&D division of RDA
were analysed for the properties of CRS-2. Samples collected by the staff of RDA
from the work sites as well as from the Emulsion Plant were also included to this
study. This analysis was done to identify the problems related to properties of CRS-
2.

41
Laboratory studies were also carried out to improve the qualities of bitumen
emulsion which were most important in construction of surface dressing. The
effect of the properties of the bitumen emulsion on surface dressing and techniques
adopted to improve bitumen emulsion (when necessary), in order to minimise the
failures identified were discussed in this chapter.

4.2 Methodology

Field investigation was carried out on roads where the surface dressing had been
completed and at on going construction sites and major common failure types were
identified.

Samples of binder were collected from the sites and from the manufacturing plant
and tested at R&D division of RDA according to the accepted standards, American
society for testing of materials, (ASTM D244, 1991) {ref; Annex-2}

Records of the bitumen emulsions for the years 1999 and 2000 were collected and
analysed for the properties, {ref; Annex-3, Test methods of bitumen emulsion}

Laboratory experiments were carried out by changing recipe formulation of

bituminous emulsion together with changing the settings such as pressure and
temperature of the manufacturing plant.

Three organic modifiers and a doping agent were tried to improve the properties of
bitumen emulsion further.

4. 3 Analysis of the Data

Special attention was given to the intrinsic properties of bitumen emulsion.

Viscosity is very important in case of surface dressing emulsions as it indicates the

flow property of the same. Bitumen emulsions with less viscosity runs off across

42
the road even due to the normal camber. Therefore the required bitumen film or the
requiied binder content does not remain on the road surface to hold the aggregate
firmly resulting higher rate of stripping of aggregate, with time. The aggregate
which are not properly bonded to the road surface will damage the windscreen
especially on cars when moving on newly laid roads. And the surface dressing
would also become poor in covering and protecting the pavement structure
underneath from inclement weather conditions.

Storage stability is another important intrinsic property of bitumen emulsion in

storing the same to maintain the homogeneity of emulsion. Failing of storage
emulsion varies viscosity from top to bottom inside the storage tank. This will
create problems in the performance of the surface dressing as well as damaging the
windscreen of the vehicles due to spraying of less viscosity portion. And also there
will be a tendency to block the nozzles of spray bar of the sprayer when the
presence of coagulated bitumen particles in storing tanks. A complement test, sieve
test was also important to get to know the presence of the coagulated particles.
Fig.4.1 shows settlement of emulsion in storage tanks.

tig: 4.1 Coagulation of bitumen particles in storage tanks

4.3.1. Storage stability

The analysis of the samples of the bitumen emulsions for storage stability is shown
below. The analysis also shows that 62% of the values satisfied the specification
limits. 28% of the values were in between 0 and -1 and 10% of the values were
above the specification limits. {Fig.4.2} Almost all the samples satisfied the

43
specification limit for the sieve test. Table 4.1 Shows the specification limits for
bitumen emulsion.

Analysis of Storage Stability

St Sta< 1 St Sta> I 0<StSta<l

Figure.4.2. Representation of the analysis of the storage stability

St. Sta. Storage stability

4.3.2 Viscosity

The analysis of results for the Saybolt Furol viscosity in seconds is shown in figure
4.3. 74% of the results were below the specification limits. Only 26% of the
results satisfied the specification limits. (ASTM 2397.1991) However according to
some other specifications the values of viscosity of CRS-2 emulsions should be in
between 150 and 300 Saybolt Furol Seconds if it is to be used as surface dressing
emulsions.

Analysis of viscosity
CRS-2

l5(Kvis<300 IO(Kvis<400 vis<100

viscosity

Figure 4.3; Representation of the analysis of viscosity

44
 Table 4.1 :Standard specification for CRS emulsions

STANDARD SPECIFICATION FOR CATIONIC EMULSIFIED ASPHALT

(ASTM Designation D 2397-85)

Requirements for Cationic Emulsified Asphalt

Type Rapid - Setting

Grade CRS - 1 CRS-2

min. max. min. max.

Test on emulsions:
Viscosity, Saybolt Furol at 77° F(25°C) - s.
Viscosity, Saybolt Furol at 122° F(50° C) - s. 20 100 100 400
Storage Stability Test 24hr. - % 1 1
Classification Test passes passes
Coating ability and water resistance:
Coating, dry aggregate
Coating, after spraying
Coating, wet aggregate
Coating, after spraying
Particle charge test positive positive
Sieve test % 0.10 0.10
Cement mixing test %
Distillation:
Oil distillate by volume of emulsion % 3 3
Residue % 60 65
Tests on residue from distillation test:
Penetration 77°F(25°C) lOOg 5s 100 250 100 250
Ductility 77°F(25°C)cm/min.cm 40 40
Solubility in trycloroethylene % 97.5 97.5

45
When considering both the above-mentioned properties of CRS-2, only 18% satisfy
the specification limits.

Therefore the analysis shows that there is an extreme necessity to improve the
bitumen emulsion. Therefore as the first step, the research commenced with
improving bitumen emulsions used in case of viscosity mainly to avoid runoff as
such reasons for less viscosity even for the samples sent from the production unit
were investigated {fig.4.4}.

Therefore the research concentrated to the production of bitumen emulsion in order

to commence the improvement of bitumen emulsion.

4.3.3 Manufacturing of bitumen emulsion

Bitumen Emulsion Stabilizer Emulsifier Acid

Colloidal mill heater water

Figure 4.4; A typical in- line emulsion plant
Bitumen is one of the major components of bitumen emulsion. Therefore the
research concentrated to the bitumen content set at the production unit. The figure
4.5 shows the variation of the bitumen content of the CRS-2 emulsion produced at
the emulsion plant.

4.4 Binder Content of the Samples of Emulsions Analysed

The graphical representation (figure 4.6) of the binder content of the samples was
analysed. 76% of the samples had the binder content in between 65 and 66

46
inclusive of the boundaries by the total weight of emulsion. Only 4% had the binder
content less than 65%. 9% of the samples had the binder content equals to 67%.
11 % of samples had binder quantity over 67%.

Binder percentage

Pb<65 65<=Pb<=66 Pb=67 67<Pb

% of bitumen

figure 4.5: Representation of binder content in the emulsion

The variation of the viscosity of emulsion at particular bitumen content was also
analysed and presented figure 4.6

Viscosity vs Residue

residue

Figure 4.6: Representation of viscosity Vs Residue

47
It was observed that there was a wide variation of viscosity for a particular binder
content while other parameters such as type and quantity of emulsifier, recipe of the
emulsion, temperature of soap solution and bitumen, temperature of emulsion at the
collection and pressure and mill rpm, mill gap remain the same.

Figure 4.7 shows that average viscosity variation of the CRS-2 emulsion against the
binder content. It can be shown that the relationship follows a polynomial equation,
2
y = 1.7386x -213.3x +6597.4 where
2
Coefficient of regression, R = 0.9894

Viscosity V s % Binder

130

62 63 64 65 66 67 68
Binder%

Figure 4.7: Representation of average viscosity Vs binder percentage

Generally viscosity variation with binder content is such that the viscosity varies at
low rate at bitumen content and higher rate at higher bitumen content. This
relationship was obtained for particular settings, such as temperature of the
manufacturing plant and also for a given formulation for emulsions. The
temperature at different stages of the manufacturing unit will effect the particle size
distribution of the emulsion even the formulation, mill type and gap are remained
fixed. In other words when the settings changes the properties of the production
will change. Further study shall be proposed to be carried out in this regard and to
find out whether there exist a series of curves for the manufacturing plant at
different temperatures. The existence of the set of curves will make easier to select
the optimum curve which gives the required properties of emulsion. The research

48
can also be further extended to find out the relationship with the mill type and gap

of particular type of mill.

Emulsions with smaller particles having narrow range will provide still better
emulsion. (Refer test results on the samples done in the Swedish laboratory)

The analysis also showed that 85% of the samples set at the binder content 65% by
the total weight of emulsion, had the viscosity 55-65 SS. 55% of the samples set at
the 6o% by the total weight of emulsion had the viscosity 65-75 SS. 61% of the
samples set at the binder content, 67% had the viscosity 75-80 SS. (Figure 4.7)

4. 5 Effect of Properties of Bitumen on Emulsion

As discussed earlier, the basic purpose of emulsion production is to combine of two

immiscible liquids (bitumen and water) to produce a serviceable binder. The
viscosity of the bitumen at the milling stage is absolutely critical, as the mill will
not satisfactorily produce globules of the desired size range when the binder does
not flow properly. In this regard temperature to which bitumen would have to be
heated would be very important. This temperature mainly depends on the type of
the bitumen used in production of emulsion. Normally the penetration is inversely
proportionate to the viscosity of bitumen. It was noted that the penetration of even a
particular grade of bitumen which is locally available often changes and hence may
affect the properties of emulsion.

Changing of penetration of a particular grade of bitumen was analysed and shown

belov-. (Figure 4.8 & Figure 4.9). The penetration of samples of bitumen labeled as
80/100-penetration grade had penetration value ranging from 70 to 115 (1/10 of
mm). Similarly the penetration of bitumen labeled as 60/70-penetration grade had
the penetration value ranging from 40 to 90 (1/10 of mm).

The mean and the standard deviation of the penetration values of 80/100 pen. grade
bitumen are 84.65 and 20.88 and that for 60/70 pen. grade bitumen are 70.3 and
11.04 respectively, {fig. 4.8 and fig.4.9}

49
 Variation of p e n .
8 0 / 1 0 0 bit.

8 ,

80 90 100 110 120

penetration

Fig.4.8 Variation of penetration of 80/100 bitumen

variation of p e n .
6 0 / 7 0 bit

40 60 80 100

pen.

Fig .4.9 Variation of penetration of 60/70 bitumen

It was decided to compare the properties of bitumen emulsion with the

corresponding base bitumen data used to produce that bitumen emulsion. This was
required because that there was a variation in the properties of same type of
emulsion produced using different batches of bitumen.

Therefore the samples of bitumen emulsion were collected with the sample of
bitumen used from the production unit RC&DC. These samples were tested at the
Central Laboratory of RDA. The summary of the results of the tests are shown
below in table 4.2.

50
The table shows that viscosity of the emulsion increases when the other conditions
remained same.

Table 4.2: Effect of penetration of raw bitumen to the properties of emulsion

Pen. R a w Type of Viscosity Storage Sieve Distillation test Tests o n residue

Bitumen emulsion ss stability test oil % residue % pen Ductility
81 CRS-1 16 0.3 0.00 2.0 64 135 118
CSS-1 18 0.2 0.00 61 185 127
CRS-1 15 0.3 0.00 2.0 65 132 116
87 CRS-1 26 0.0 0.00 2.5 65 140 134
CRS-1 20 0.3 0.00 1.5 60 133 126
90 CRS-1 20 0.0 0.00 2.5 60 105 125
CRS-1 22 0.1 0.00 1.5 59 97 145
CRS-1 21 1.0 0.00 1.5 56 123 150
CRS-1 21 0.0 0.00 2.5 58 105 120
102 CRS-1 30 0.2 0.00 2.0 63 105 95
CRS-1 28 1.0 0.00 2.0 61 110 132
CRS-1 25 1.0 0.00 2.0 60 163 94
CRS-1 24 0.6 0.00 2.5 60 96 140
CRS-1 29 0.5 0.00 2.5 63 100 122
105 CRS-1 36 1.0 0.00 1.5 64 103 131

The study also revealed that changing the temperature/settings to suit the viscosity
of bitumen at the emulsion plant will provide a feasible solution to maintain the
range of the desired particle size of bitumen globules and hence better properties of
bitumen emulsion. A further study is needed to establish a calibration chart of
temperature/settings of the manufacturing unit and the penetration of bitumen,
provided that the penetration of bitumen is known. For this the production unit of
bitumen emulsion must have the facility to test the penetration of bitumen prior to
use.

However there are other parameters which will affect the particle size distribution
of emulsion and will be discussed later in this chapter.

Three samples were tested at the Emulsion laboratory, Scan Road at Nacka,
Stockhorm (Sweden) for the particle size distribution of bitumen emulsion using the
Granulometer HR 850 the results are tabulated below, {table 4.3 & 4.4} The

51
particle size distribution is unable to test in the RDA laboratory. Test reports are
attached in the annex-10

CaC12 is used to improve the stability of bitumen emulsion. The table 4.3 shows
that the particle sizes of the three different bitumen emulsion samples done using
the Granulometer HR850. The table explains that the mean size of the bitumen
globules has become smaller with the presence of CaC12. And the range of the
particle size distribution also has become narrow. This indicates that adding of
CaC12 will improve the particle size of bitumen emulsion. For instance storage
stability improves as the range of the particle size distribution becomes narrow and
viscosity increases because the mean size of the particles is smaller, compared the
non-modified samples. The same table also shows that the modified emulsion with
chemical-1 has the best particle size distribution and the least mean size of particles
distribution and the least mean size of particles, such that the chemical -1 (Di-
Amine OLB) will be a better modifier than to CaC12.

Table 4.3: Particle size distribution of bitumen emulsion

Item type Median size Dia. for 10% Dia. for 90%
microns microns microns
1 No each 4.54 0.75 16.38
2 With each 3.88 0.75 15.76
3 With chem.l 3.04 0.58 9.58
Note: Dia. - Diameter

The above table shows that the medium size of the particles modified with chemical
1 is 3.04 microns which is the least of the three and the narrowest range of the
particle size distribution. This indirectly shows that there is an improvement in
emulsion specially in case of viscosity. (Table 4.4)

Table 4.4: Viscosity of the samples

Item No each With each With chem.l
Viscosity ss 12 14 20

52
The research scheduled to study the factors which effects the properties are as
mentioned below.
1. The formulation variables that are dealing with the ingredient nature, (type
of emulsifier, type of bitumen)
2. The composition variables that describe the different ingredients
proportions.
3. The mechanical aspects of mixing device the energy input and its duration.

1. The research study about the emulsions was initiated by varying the type of
emulsifiers. The emulsifiers tested for CRS emulsions were redicote E-965, Asifer
Nl 00L and Dinoram SL. These results are shown in table 4.5 & table 4.6.

Table 4.5: Effect of type of emulsifier

Type Viscosity Storage Sieve Residue Oil Tests on residue

SS Stability Test % % % pen. Duct.
Dinoram 16 0.4 0.3 62 3.0 110 145
SL
Asifer 21 -0.2 0.0 62 3.0 109 130
N100L
Redicote 70 -1.2 0.08 63 2.5 108 89
E965.

Formulation of above emulsions were done based on the so-called optimum

formulation which were proposed by the manufacturer of the particular emulsifier.
The optimum formulations are as shown below

Table 4.6; optimum formulations used as manufacturers guide

Emulsifier Bitumen Kerosene Emul. % Cacl 2

pH Water %
Dinoram 61 2.0 0.16 0.1 2.5 36.74
SL
Asifer 61 0.16 0.1 2.0 38.74
N100L
Redicote 61 1.0 0.25 0.1 2.5 37.65
E965

53
2. The study was carried out by changing the composition variables as the second
step. This step includes the changing of the proportion of an emulsifier, by
changing the proportions of solvent, kerosene and the proportion of bitumen. The
values are tabulated as below.

The table 4.7 shows that the properties of bitumen emulsion increases with the % of
emulsifier. Bitumen emulsions having higher quantity of emulsifier makes stable
emulsions. However creaming of emulsions may possible because of high viscosity
water media.

Table 4.7; Effect of emulsifier on CSS emulsions

Emul. % Vis. St.Sta. Sieve Residue Oil Tests on residue

emul. SS % Test % % % Pen Duct.
Redi. 0.8 20.00 0.70 0.00 61.00 2.0 99 68
4875
Redi. 1.0 22.00 0.80 0.00 61.00 2.0 102 120
4875
Redi. 1.2 25.00 0.80 0.00 61.00 2.0 100 101
4875
Em 26 0.7 16.00 0.10 0.00 61.00 2.0 125 55

Em 26 0.9 16.00 0.30 0.02 61.00 2.0 127 89

Em 26 1.0 17.00 0.70 0.01 60.00 2.0 124 96

Em 26 1.2 18.00 0.10 0.00 61.00 2.0 134 90

The table 4.8 shows that the viscosity of bitumen emulsion increases with
increasing the kerosene content. This can be explained when the kerosene content
increases it is easier to break down bitumen into finer globules than when there is
no kerosene. It was also observed that penetration of the residue (base binder) also

54
increases with the kerosene content because bitumen dissolves in kerosene and
hence becomes softer. However care should be taken not to increase kerosene
much because it takes more time to cure at the field.

Table 4.8; Effect of the solvent on the properties of emulsion CRS-1

Solvent Viscosity Storage Sieve Residue Oil Test on residue

% stability Test % % % Pen. Duct.

0.0 21 0.2 0.0 65 1.0 96 115

2.5 43 0.2 0.02 65 2.0 165 80

Formulation tried
Formula-1
Emulsifier Asifer N100L
Quantity 0.16%
Cacl 2 0.1%
pH 2.5
Settings of the emulsion plant were not changed.

Table 4.9: Effect of quantity of bitumen on the properties of emulsion

Bitumen Viscosity Storage Sieve Residue Oil Test on residue

% stability Test % % % Pen. Duct.
60 21 0.0 0.00 61 2.0 133 119
65 30 -0.6 0.00 64 2.0 115 133
68 43 0.4 0.12 67 2.0 163 80

The table 4.9 also shows that the normal behavior of increasing viscosity of
emulsions at high bitumen content.

Formula -1 (mentioned above) was used and the settings of the plant were not
changed.

55
3. The study of changing the mill type was not possible because only one laboratory
mill was available in the laboratory. Therefore the study confined only to establish
the mill rpm at RC&DC emulsion plant which produces emulsions having almost
similar qualities as emulsions produced at the laboratory

Emulsion produced using laboratory emulsion plant at 9000rpm, 145°C bitumen

temperature and 80°C emulsion was very similar to 5060rpm, 145°C bitumen
temperature and 80°C emulsion temperature at RC&DC production unit. The
research also extended to study the variation of bitumen emulsion with the mill
rpm.

Table 4.10: Effect of mill rpm on the properties of emulsion

Item mill r p m water p u m p Viscosity Storage Sieve Distillation test

rpm SS Stability test oil % Residue %
1 5000 110 16 0.4 0.05 2.5 62.2
2 5060 109 20 0.1 0.00 2.5 61.5
3 5500 110 18 0.2 0.01 2.5 63.4
4 6000 110 19 0.3 0.00 2.5 62.0
5 6500 110 19 0.2 0.00 2.5 62.0
6 7000 109 21 0.1 0.01 2.5 62.0
7 7000 109 21 0.5 0.02 2.5 64.0
8 9030 110 23 0.3 0.00 2.5 61.4
9 9030 110 24 0.2 0.00 2.5 61.9
10 9040 110 24 0.2 0.00 2.5 61.5
11 9000 110 25 0.1 0.01 2.5 62.0
12 9000 110 24 0.2 0.01 2.5 61.4
13 10,000 109 25 0.2 0.02 2.5 61.5
14 11,000 110 19 0.3 0.0 2.0 61.6

The table 4.10 & figure 4.10 show the variation of the viscosity with mill speed,
rpm. The viscosity changes with rpm such that increases with increasing rpm. The
study also further investigated that the variation follows a certain polynomial
distribution having the following equation. The regression coefficient is .9685
which indicates that a good co-relation between the two variables.

2
y = 5E-07x - 0.0047x + 28.694

2
R = 0.9685

56
For particular settings there exists a curve for the variation of viscosity with the mill
speed. This may due to the variation of particle size distribution which is one of the
main governing factor for viscosity. There is a possibility to have a series of curves
in this regard for different settings which has the main advantage of selecting the
particular mill speed and the settings to attain the viscosity required.

Viscosity Vs Mill rpm

CRS-1 Emulsion

0 • • . • - . ,
4000 5000 6000 7000 8000 9000 10000 11000
mill rpm

Figure 4.10: variation of viscosity Vs mill rpm

Changing the mill gap is not possible in all the types of production units. However
rings of different gauges and the same diameter are available in the laboratory
emulsion plant. These rings can be inserted in the gap in order to reduce the mill
gap. Mill gap is reduced more if a ring of higher gauge is inserted.

4.6 Conclusions

Properties of bitumen emulsions can be improved by changing the above-mentioned

factors. However these prime factors may not be continuously increased or
decreased until the required improvement is gained. For instance bitumen content
should not be increased above 70% because the result will be an inverted emulsion.
Therefore the study intended to research for the modifiers.

57